J. Agric. F O O ~ Chem. m
such as the fiberoptics luminoscope skin monitor. No.5-M0P, 484-20-8; 8-MoP, 298-81-7; 66-97-7.
LITERATURE CITED Beaumont, P. C.; Parsons, B. J.; Navaratnan, S.; Phillips, G. 0. Allen, J. C. Biochim. Biophys. Acta 1980, 608, 259. Beier, R. C. J. Liq. Chromatogr. 1985,8, 1923. Beier, R. C.; Oertli, E. H. Phytochemistry 1983, 22, 2595. Beier, R. C.; Ivie, W. G.; Oertli, E. H.; Holt, D. L. Food Chem. Toxicol. 1983. 21. 163. Berkley, S. F.; Hightower, A. W.; Beier, R. C.; Fleming, D. W.; Brokopp, C. D.; Ivie, G. W.; Broome, C. V. Ann. Int. Med. 1986, 105, 351. Hearst, J. E. Ann. Rev. Biophys. Bioeng. 1981, 10, 69. Mantulin, W. W.; Song, P. S. J. Am. Chem. SOC.1973,95,5122.
a , 36,337-340
337
Oertli, E. H.; Beier, R. C.; Ivie, G. W.; Rowe, L. D. Phytochemistry 1984, 23, 439. Parrish, J. A.; Fitzpatrick, T. B.;Tannenbaum, La;Pathak, M. A. New End. J. Med. 1974. 1207. Scott, B. R.; fathak, M. A.; Mohn, G. R. Mutat. Res. 1976, 29. Seligman, P. J.; Mathias, C. G. T.; O'Malley, M. A.; Beier, R. C.; Fehrs, L. J. Arch. Dermatol. 1987, 123, 1478. Vo-Dinh, Am. Ind. Hyg. Assoc. J. 1987, 48, 594. Vo-Dinh, T.; Gammage, R. B. Am. Ind. Hyg. Assoc. J. 1981, 42, 112. Vo-Dinh, T.; White, D. A. Anal. Chem. 1986,58, 1128. Received for review September 12,1986. Accepted September 17,1987. Research sponsored by the Office of Health and Environmental Research, U.S. Department of Energy, under Contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc.
Analysis of Ergopeptine Alkaloids in Endophyte-Infected Tall Fescue Shelly G . Yates* and Richard G . Powell
A method has been developed for analysis of ergopeptine alkaloids in tall fescue by means of highperformance liquid chromatography (HPLC) with fluorescence detection. Ergopeptine alkaloids in methanol extracts of seed were routinely analyzed in an acetonitrile-pH 7.6 buffer gradient with dihydroergocristine as an internal standard. Measured recovery of ergovaline added at the 1.5 pg/g level was 98%, and the lower limit of detection was 150 pg. Ergopeptine alkaloids, primarily ergovaline, are presumed contributors to animal toxicity of tall fescue (Festuca arundinacea) infected with Acremonium coenophialum. The method complements tandem mass spectrometry (MS-MS) in that HPLC readily distinguishes physiologically active "ine" and inactive "inine" epimers.
Tall fescue (Festuca arundinacea Schreb.) was brought to the new world by early settlers and existed in relative obscurity for many years. A pure stand of a naturalized variety was observed in Kentucky in 1931, and this variety, KY-31, was released commercially in 1943 (Ball, 1984) and recommended for use by the Department of Agriculture. Tall fescue gained rapidly in popularity (Cowan, 1956),and recently forage scientists estimated that 35 million acres are grown in the southeastern United States (Ball, 1984). Much of the time, tall fescue compares favorably with other pasture grasses with respect to daily weight gain of cattle on pasture. Under heat stress during summer months, however, clinical signs of fescue toxicity, including elevated temperature, poor weight gain, problems with reproduction, and reduced lactation (Bush et al., 1979), are often observed. This phenomenon, referred to as summer slump or summer syndrome, has become accepted as normal by livestock producers who rely primarily on fescue forage. During periods of stress due to cold in winter, another syndrome known as fescue foot can become important (Bush et al., 1979). Characteristic signs of fescue foot are loss of weight, rough hair coat, arched back, trembling, and gangrene of the extremities in severe cases. Despite these reports and estimated economic losses of $50 to $200 million annually (Siege1 et al., 1984), tall fescue is favored as a cool season pasture for year-around grazing Northern Regional Research Center, U.S.Department of Agriculture-Agricultural Research Service, Peoria, Illinois 61604.
systems. The problems with fescue can be largely avoided by proper management practices. Full understanding of the causes of these syndromes remains elusive due to lack of understanding of all contributing factors and the need for rapid and sensitive analytical procedures. Symptoms of fescue toxicity often mimic those of ergotism; however, ergot (Claviceps purpurea) does not appear to be the major problem in fescue pastures. In recent years it has been recognized that 95% or more of the tall fescue grown in the United States is infected, to varying degrees, with Acremonium coenophialum and that the degree of toxicity of fescue forage is related to the degree of infection by this endophytic fungus. A. coenophialum has been demonstrated to produce ergopeptine alkaloids, notably ergovaline (Porter et al., 1981; Yates et al., 1985), and, though not present in endophyte-free fescue, ergovaline appears to be a normal constituent of endophyte-infected fescue leaf blades and sheaths at levels up to 2.8 ppm (Lyons et al., 1986). The presence of ergovaline and other ergot alkaloids in fescue at reported levels is of concern due to their wide range of potent physiological activities (Berde and Schild, 1978). In addition, concentrations of these alkaloids in forage may fluctuate widely due t o seasonal or environmental changes (Lyons et al., 1986), unrecognized analogues may be present, and additive or synergistic effects involving other types of compounds may be complicating factors. This report describes a rapid, sensitive, and reasonably quantitative method for analysis of ergot alkaloids in tall fescue. The procedure is a modification of an HPLC method for analysis of ergot alkaloids in flour (Scott and
This artlcle not subject to US. Copyrlght. Published 1988 by the American Chemical Society
338 J. Agric. Food Chem., Vol. 36, No. 2, 1988
Lawrence, 1980) programmed for maximum separation of those compounds most important in tall fescue studies (ergonovine, ergovaline, ergosine, ergotamine). The use of dihydroergocristine as an internal standard allows for increased sensitivity and confidence levels.
Ergopeptine Alkaloids Ergotamine Ergosine 8-Ergosine Ergovaline Ergostine Ergoptlne 8-Ergoptine Ergonine Ergocrystine a-Ergocryptine 8-Ergocryptine Ergocornine
R’
A‘
CH3 CH3 CH3 CH, C2H6 C,& CzH6 CzH6 i-Pr I-Pr I-Pr I-Pr
PhCHz I-BU sec-Bu i-Pr PhCH2 i-Bu sec-Bu i-Pr PhCH2 i-Bu sec-Bu i-Pr
Yates and Powell
25.0 X for ergotamine, and 8.2 X for ergonovine. For routine analyses, an average response factor (R3 of 23.9 x was used for all ergopeptide alkaloids. Quantities of individual alkaloids were calculated by the formula pg alkaloid/g sample = Rf X area alkaloid/area dihydroergocristine X pg dihydroergocristine/g sample
Solvents and Gradient Program. Solvents consisted of (a) HPLC-grade acetonitrile and (b) 0.1 N ammonium acetate buffer (pH 7.6), refrigerated when not in use. Particulate matter and dissolved gases were removed by vacuum filtration using a 0.45-pm nylon 66 membrane. The gradient program, used for quantitation by the peak height method with a flow rate of 0.8 mL/min, was as follows: time, min 0.0
0.1 40.0
75.0 100.0
The gradient program used for quantitation by the internal standard method, with a flow rate of 0.8 mL/min, was as follows: time, min
0
0.0
0.1 35.0 70.0
85.0
Ergonovine EXPERIMENTAL SECTION
Instruments. Programmed-gradient HPLC was performed with a Beckman 421 controller and llOA pumps, and detection was accomplished by a Kratos Spectroflow 980 fluorescence detector (excitation at 310 nm, 360-nmlong pass filter, 1-s rise time, range 0.025 pA full scale). Detector response was integrated by a computer. HPLC was conducted on DuPont Zorbax columns (4.6 X 250 mm, 5 pm ODs, or 4.6 X 250 mm, 5 pm C-8) fitted with a 20-pm Zorbax ODS guard column and a 4.6 X 250 mm precolumn filled with 200/425-mesh Adsorbosil (Alltech). Standard Alkaloids. Ergotamine, a-ergocryptine, and ergonovine were purchased from Sigma Chemical Co., dihydroergocristinemethanesulfonate was purchased from Atomergic Chemetals Corp., and ergovaline was a gift from Dr. George Rottinghaus, University of Missouri. Other standard alkaloids were gifts of Sandoz Ltd., Basel, Switzerland. TLC analyses, molecular extinction coefficients, and MS and NMR spectra of all alkaloids indicated that they were correctly identified and were of sufficient purity to be used as quantitative standards (Pollock and Stevens, 1965; Yates et al., unpublished Data). Dihydroergocristine, added at 100-fold the expected alkaloid concentration, or 0.4 mg/g of sample (plant material or feed ration), was used as an internal standard. In earlier experiments quantitation was achieved by measuring peak height, and nanograms of alkaloid per millimeter of peak height were established for pure alkaloids by repetitive analyses. Response factors = nanograms of alkaloid/nanogram of dihydroergocristine X area dihydroergocristine/area alkaloid, determined by repetitive analysis, were 22.7 X for ergovaline, 23.9 X for ergosine,
event 35% CHBCN 50% CH3CN in 30 min 100% CH&N in 15 min 35% CH3CN in 5 min end
event 35% CH3CN 50% CH3CN in 30 min 100% CH&N in 15 min 35% CH3CN in 5 min end
During periods when the column was not in use, it was cleaned with a flow rate of 0.8 mL/min with the following buffer flushing program: time, min 0.0
15.0 65.0
120.0 150.0
event 35% CH3CN, 65% HPLC water 100% HPLC water in 10 min 100% CH&N in 15 min 85% CH3CN, 15% HPLC water in 10 min hold and stop flow
Sample Preparation and Recovery Experiments. Various recovery procedures were compared including exhaustive Soxhlet extraction with hexane followed by methanol (agitation), sequential extraction at room temperature with a solvent series (3X each; hexane, ethanol, and 50% aqueous ethanol), and simple extraction (3X) with either CHC13or MeOH alone. The preferred method for HPLC analysis of ergot alkaloids was to extract samples of endophyte-free or infected seed either alone or mixed with Purina rat chow 1:1, 3X at room temperature with MeOH (10 mL of solvent/g of sample). Preparation of combined extracts for analysis involved filtering through Whatman No. 54 paper, reduction to an appropriate volume at 25 “C, and final filtration through a 0.45-pm membrane. Internal standard, when used, was added to the initial sample (0.4 mg/g) prior to extraction. All operations were conducted in subdued light. Endophyte-free and endophyte-infected seed were obtained through the courtesy of the University of Kentucky or purchased from Lambert Seed Co., Camden, AL. The latter seed was certified as 100% endophyte-infected. RESULTS AND DISCUSSION
The relatively simple lysergic acid amide, ergonovine, and the 12 common ergot peptide alkaloids, along with their corresponding C-8 epimers, were all examined in this study. Chromatography of reference standards revealed
J. Agric. Food Chem., Vol. 36, No. 2, 1988
Ergopeptine Alkaloids in Endophyte-Infected Fescue
Table I. Order of Elution of Ergonovine, Ergopeptine Alkaloids, and Their Epimers on Reversed-Phase Columnsa alkaloid MW 441 ergonovine 441 ergonovinine 533 ergovaline 547, 547b ergosine & P-ergosine 547 ergonine 581 ergotamine 561, 561b ergoptine & P-ergoptine 561, 595b ergocornine & ergostine ergovalinine 533 a-ergocryptine 575 575, 60g6 6-ergocryptine & ergocristine 547, 547b ergosinine & 8-ergosinine 547 ergoninine 581 ergotaminine ergoptinine, 6-ergoptinine, ergocorninine, and 561, 561, 561,b611 dihydroergocristinec ergostinine 595 575, 5756 a-ergocryptinine & 0-ergocryptinine ergocristinine 609
"In order of increasing retention time. bCompounds are not separated when gradient program described in the Experimental Section is used. Semisynthetic internal standard. b
Table 11. Change in Ergovaline Response Factor with Time day response factor 9 ' 0 change 0 22.7 x 10-3 0 1 23.6 x 10-3 4 2 24.5 X 8 7 25.7 x 10-3 13 18 27.0 x 10-3 19
the order of elution as shown in Table I. Elution time increased with molecular weight with two notable exceptions; those alkaloids having benzyl substituents (ergotamine, ergostine, ergocristine) eluted sooner than predicted based on their aliphatic counterparts having similar molecular weights. Those compounds considered to be naturally occurring all eluted earlier than the corresponding C-8 epimers. These C-8 epimeric compounds, assumed to be artifacts of isolation, exhibit markedly reduced physiological responses and are generally less soluble than their naturally occurring counterparts (Berde and Schild, 1978). Dihydroergocristine proved to be an acceptable internal standard as it eluted in a region where interference from other fluorescent peaks was tolerable; it does not occur naturally, and it does not epimerize at C-8. Dihydroergocristine eluted later than the alkaloids of interest, and its fluorescence response proved to be much less than the response of the normal unsaturated alkaloids. Fluorescence response, expressed as millimeters of peak height/nanogram of alkaloid, of all the common ergopeptide alkaloids was similar on RP-8 and RP-18 columns (6.4 mm/ng, SD = 1.0 and 12.2 mm/ng, SD = 1.1, respectively); the response of ergonovine (30 mm/ng) was more intense. Fluorescence response factors for pure compounds were determined as listed in the Experimental Section. Response factors were determined immediately upon dissolving the samples in methanol as the compounds are somewhat unstable in solution. Ergovaline, kept in methanol at room temperature for 6 h during analysis and then stored at -18 "C, showed a continuous drop in response over an 18-day period (Table 11). Isomerization of ergovaline, to ergovalinine, occurred readily and was observed to progress in methanol solution until an equilibrium mixture (near 1:l) was obtained. Loss of fluorescence response was assumed to be due to reaction of the alkaloids with solvent and exposure to light (Berde and Schild, 1978).
339
I' I
I1
1
I
I
I
I
1
0
10
20
30
40
I
I
50 60 Time, min
I
I
70
80
90
Figure 1. HPLC separation of standard alkaloids important to tall fescue research. Peaks: 1,ergonovine; 2, ergonovinine; 3, ergovaline; 4, ergosine; 5, ergotamine; 6, ergovalinine; 7, ergosinine; 8, ergotaminine. Table 111. Precision of HPLC Analysis of Ergot-like Alkaloids in Endophyte-Infected Tall Fescue Seed rg of alkaloid/g of seed" alkaloid mean SD unknownb 1.23 0.10 unknown 0.95 0.07 ergovaline 2.89 0.22 unknown 0.15 0.02 ergosine 0.80 0.31 ergotamine 0.21 0.12 ergovalinine 2.51 0.28 ergosinine 0.47 0.12 ergotaminine 0.32 0.08
."Injections of 4 or 5 rL representing 1600 or 2000 r g of seed. Mean of seven analyses. Peak contains both unknown and ergonovine. This pair is partially resolved if each is present in approximately equal amounts. SD = standard deviation. Table IV. Recovery of Alkaloid Standards kglg of seed endogenous added ergovaline, ergovalinine 5.40 1.43 ergosine, ergosinine 1.27 0.23 ergotamine, ergotaminine 0.53 0.18
found 6.67 2.20 1.42
A study of the extraction process revealed that when rat rations spiked with ergonovine, ergocryptine,or ergotamine were extracted, 80-87% of the total extractable alkaloid was found in the first extract and that 0-5% was found in the third extract. Removal of ergot alkaloids from ground seed was more difficult; 61-73% of the total extractable alkaloid was found in the first extract, and 7-970 of the total alkaloid was in the third extract. A typical HPLC run for standards is shown in Figure 1. This may be compared to HPLC runs of extracts equivalent to 1400 pg of endophyte-infected seed (Figure 2A) and to 1400 pg of endophyte-free seed (Figure 2B). A typical HPLC record of endophyte-infected seed, with dihydroergocristine added as an internal standard and a slightly different program gradient, is given in Figure 3. Peaks eluting in the first 5 min and peaks eluting after 53 min occur in both endophyte-free and endophyte-infected samples. The former peaks were unidentified compounds extracted from the seed, and the latter peaks were due to unidentified impurities in the solvent. The precision obtained with seven representative samples from the same lot of fescue seed is given in Table I11 (samples are representative of 1600 or 2000 wg of seed). Recoveries of ergovaline, ergosine, and ergotamine added
340
J. Agric.
Food Chem., Vol. 36, No. 2, 1988
Yates and Powell
Progress in understanding the problem of fescue toxicity has been hampered by lack of knowledge concerning what chemical factors are responsible for the observed problems, general unavailability of rapid and accurate analytical methods for known contributing factors that are present only at the ppm level, and unavailability of an appropriate small animal model or other bioassay. The method described herein was developed for rapid and accurate analysis of ergopeptide alkaloids and may be used to analyze fescue samples or to monitor feeding trails. The necessary equipment for HPLC analysis is available in most chemical laboratories and is relatively inexpensive compared to mass spectrometric techniques. ACKNOWLEDGMENT 0
30
20
10
50
40
70
60
80
90
Time, min
Figure 2. HPLC separation of an ethanol extract representative of 1400 fig of endophyte-infected (A) and endophyte-free (B) tall fescue seed. Peaks: 1, unknown plus ergonovine; 2, unknown; 3, ergovaline; 4, ergosine; 5, ergotamine; 6, ergovalinine.
t
9
~
10
-*
20
-1
1.
.
30
40
5D
._ .
60
I--
70
-A_,
80
Time niin
Figure 3. HPLC separation of ergopeptine alkaloids of endophyteinfected Ky-31 tall fescue seed with added internal standard, dihydroergocristine. Separation accomplished on a RP-18 analytical column with a gradient program with pH 7.6 ammonium acetate buffer and acetonitrile. Four microliters of methanol extract representing 1600 pg of seed was injected. Peaks: 1, unknown plus ergonovine (1.09 Wg/g); 2, unknown (1.09 pg/g); 3, ergovaline (2.77 pg/g); 4, unknown (0.13 pg/g); 5, ergosine (0.43 pg/g); 6, ergotamine (0.10 pg/g); 7, ergovalinine (2.19 pg/g); 8, ergosinine (0.39 pg/g); 9, ergotaminine (0.24 pg/g); 10, dihydroergocristine.
to endophyte-infected seed, extracted into methanol, and analyzed with no further purification are indicated in Table IV. Recoveries of the latter two alkaloids were somewhat variable and appeared to be related to contamination of the seed samples with small amounts of ergot sclerotia. Careful examination of the seed lot revealed the presence of trace amounts of ergot sclerotia. Ergosine and ergotamine are characteristic alkaloids of ergot sclerotia (Porter et al., 1987), and their concentrations varied markedly from sample to sample. Recovery of ergovaline was more consistent as it is the predominant ergopeptide of the fescue endophyte and is only a minor constituent of ergot. Results of HPLC analysis for ergovaline were in reasonable agreement with those obtained by tandem mass spectrometry (Plattner et al., 1983);detailed comparisons however, were not made.
We thank Dr. Lowell Bush and Dr. Roger Hemken, University of Kentucky, for tall fescue seed samples; Dr. George Rottinghaus, University of Missouri, for ergovaline; and Dr. A. Lindenmann and H. Friedli, Sandoz Ltd., for 1 2 ergopeptine alkaloids. Registry No. Ergovaline, 2873-38-3; ergosine, 561-94-4; ergotamine, 113-15-5; ergovalinine, 3263-56-7; ergosinine, 596-88-3; ergotaminine, 639-81-6.
LITERATURE CITED Ball, D. M. “An Overview of Fescue Toxicity Research”. AgriPractice 1984, 5, 31-36. Berde, B., Schild, H. O., Eds. Ergot Alkaloids and Related Compounds; Springer-Verlag: New York, 1978. Bush, L.; Boling, J.; Yates, S. “Animal Disorders”. In Tall Fescue Monograph; Bush, L., Buckner, R., Eds.; America1 Agronomy Society: 1979; pp 247-264. Cowan, J. R. “Tall Fescue”. In Advances in Agronomy; Academic: New York, 1956; Vol. 8, pp 283-320. Lyons, P. C.; Plattner, R. D.; Bacon, C. W. “Occurrenceof Peptide and Clavine Ergot Alkaloids in Tall Fescue Grass”. Science (Washington, D.C.) 1986,232, 487-489. Plattner, R. D.; Yates, S. G.; Porter, J. K. “Quadrupole Mass Spectrometry/Mass Spectrometry of Ergot Cyclol Alkaloids”. J . Agric. Food Chem. 1983,31, 785-789. Pollock, J. R. A., Stevens, R., Eds. Dictionary of Organic Compounds; Oxford University: New York, 1965. Porter, J. K.; Bacon, C. W.; Robbins, J. D.; Betowski, D. “Ergot Alkaloid Identification in Clavicipitaceae Systemic Fungi of Pasture Grasses”. J. Agric. Food Chem. 1981, 29, 653-657. Porter, J. K.; Bacon, C. W.; Plattner, R. D.; Arrendale, F. F. “Ergot Peptide Alkaloid Spectra of Claviceps-Infected Tall Fescue, Wheat, and Barley”. J. Agric. Food Chem. 1987,35,359-361. Scott, P. M.; Lawrence, G. A. “Analysis of Ergot Alkaloids in Flour”. J . Agric. Food Chem. 1980, 28, 1258-1261. Siegel, M. R.; Johnson, M. C.; Varney, D. R.; Nesmith, W. C.; Buckner, R. C. Bush, L. P.; Burrus, P. B.; Jones, T. A,; Boling, J. A. “A Fungal Endophyte in Tall Fescue: Incidence and Dissemination”. Phytopathology 1984, 74, 932-936. Yates, S. G.; Plattner, R. D.; Garner, G. B. “Detection of Ergopeptine Alkaloids in Endophyte-Infected, Toxic Ky-31 Tall Fescue by Mass Spectrometry/Mass Spectrometry”. J. Agric. Food Chem. 1985,33, 719-722. Yates, S. G.; Powell, R. D.; Weisleder, D.; Plattner, R. D., Northern Regional Research Center, Peoria, IL 61604, unpublished UV, NMR. and MS data. Received for review August 19,1987. Accepted December 9,1987. Mention of firm names or trade products does not imply that they are endorsed or recommended by the U.S. Department of Agriculture over other firms or similar products not mentioned. Presented at the 194th National Meeting of the American Chemical Society, New Orleans, LA, Aug 30-Sept 4, 1987.